It was a watershed moment in the wonderful world of cryptocurrencies, ICOs, and blockchain technology projects. Prodeum — which promised to revolutionize the fruit and vegetable industry — replaced its website, post-ICO, with a white screen that contained just one word.

“Penis.”

It is unclear who perpetrated the scam. While the company looked like a legitimate blockchain startup based in Lithuania, there are various threads that suggest it was run by an individual in Colombia. And while they only got away with $22,000 worth of ETH (significantly more than the $11 claimed in other articles on the subject), other scams have been more fruitful.

Confido managed to walk away with over $374,000 in November 2017.

Today, WatermelonBlock — an AI-powered investment and trading platform for cryptocurrency investors and traders — has announced it is integrating with IBM Watson’s AI computing platform to provide investors with real-time insights and detailed analysis that help identify scams like Prodeum and Confido.

WatermelonBlock takes keywords, hashtags, and metadata terms relating to cryptocurrencies and ICOs from a wide variety of social and traditional media APIs. IBM Watson then measures this data for sentiment. It also weighs each message author individually, according to their social influence and reach.

WatermelonBlock then uses its algorithms to compute a percentage and index score for each network, known as the MelonScore.

Prodeum was hard to detect because it looked like a regular ICO, so how does the MelonScore help with carefully orchestrated scams?

“WatermelonBlock is designed with retail consumers in mind,” WatermelonBlock cofounder Elliot Rothfield told me. “This scam is a product of a developing market. During an era of ferment, rapid growth and changing standards make discussion difficult. By combining sentiment analysis (the voice of the people) with weighted influencer sentiment (the voice of the knowledgeable), users can circumvent being entangled in a ‘Penisgate’ controversy.”

In addition to helping investors avoid scams, WatermelonBlock is a useful source of intelligence about candidates for the ICO market, the majority of which offer legitimate projects.

By continually scanning the internet for sentiment data and analyzing both tone and author credibility, the AI-powered market predictions can even help investors spot potential winners. Whenever sentiment changes around a particular cryptocurrency, the system notifies users in real time, giving them the opportunity to anticipate market fluctuations and inform appropriate action.

That said, the MelonScore is not a predictor of future market value.

“The MelonScore is unique in that it will represent the sentiment of the masses with respect to cryptocurrency,” Rothfield said. “AI is used to create a ranking system unique to WatermelonBlock, built on big datasets gathered from social media, blogs, news, microsites, and other public forums.”

The use of IBM Watson for AI-powered analysis is just the beginning for WatermelonBlock.

“WatermelonBlock is not just a single application, but a suite of AI analysis tools,” Rothfield said. “WatermelonAnalytics will be introduced soon as a small business sentiment analyzer. [It] will allow businesses to search, analyze, and compare individual phrases, hashtags, or direct URLs to harness industry-specific insights. Users will be able to create their own private index, allowing them to track not only the sentiment of a brand, but the sentiment of certain phrases, products, and releases. WatermelonBlock’s AI and proprietary algorithms are versatile and will be used in many different products and industries. Stay tuned for WatermelonMusic, too.”